Add debug logging to track data loading on HuggingFace Space

- Add data integrity check in setup_data.py
- Add sample data logging in simple_data_loader.py
- This will help identify where the zero scores are coming from

Co-authored-by: openhands <[email protected]>

setup_data.py

@@ -69,6 +69,15 @@ def fetch_data_from_github():
         shutil.copytree(results_source, target_results)
         
         print(f"Successfully fetched data from GitHub. Files: {list(target_dir.glob('*'))}")
+        
+        # Verify data integrity by checking a sample agent
+        sample_agents = list(target_results.glob("*/scores.json"))
+        if sample_agents:
+            import json
+            with open(sample_agents[0]) as f:
+                sample_data = json.load(f)
+                print(f"Sample data from {sample_agents[0].parent.name}: {sample_data[0] if sample_data else 'EMPTY'}")
+        
         return True
         
     except subprocess.TimeoutExpired:

simple_data_loader.py

@@ -206,6 +206,13 @@ class SimpleLeaderboardViewer:
                 # Simple mapping: each tag maps to itself
                 self.tag_map = {tag: [tag] for tag in sorted(all_tags)}
             
+            # DEBUG: Print sample of loaded data
+            print(f"[DATA_LOADER] Loaded {len(transformed_df)} agents")
+            if len(transformed_df) > 0:
+                sample_cols = ['agent_name', 'overall_score', 'overall_cost']
+                available_cols = [c for c in sample_cols if c in transformed_df.columns]
+                print(f"[DATA_LOADER] Sample row: {transformed_df[available_cols].iloc[0].to_dict()}")
+            
             return transformed_df, self.tag_map
         except Exception as e:
             import traceback